DOI: https://doi.org/10.21203/rs.3.rs-1845876/v1
Background
Maintenance of low central venous pressure (CVP) during hepatic resection to minimize intraoperative blood loss requires compensative administration of large amounts of fluid after its completion. However, this fluid challenge may influence on the postoperative outcomes of patients with left ventricular diastolic dysfunction (LVDD) who cannot tolerate volume adjustment.
Methods
A total of 206 patients with who underwent hepatic resection between March 2015 and February 2021 were evaluated. LVDD was defined according to the American Society of Echocardiography and the European Association of Cardiovascular Imaging 2016 recommendations as LVDD (group A, n = 39), or normal LV diastolic function and indeterminate decision (group B, n = 153).
Results
CVP at the end of surgery was significantly higher in group A than in group B (6.35 ± 3.45 vs. 5.12 ± 2.80, P = 0.025). Postoperative acute kidney injury (AKI, 10.3% vs. 1.3%, P = 0.004) and pleural effusion or edema (51.3% vs. 30.1%, P = 0.013) were more common in group A than in group B. Further, creatinine levels from postoperative day 1 to day 7 were significantly higher and daily urine outputs at postoperative day 1 (P = 0.038) and day 2 (P = 0.025) were significantly lower in group A than in group B. LVDD was the only significant risk factor for postoperative AKI after hepatic resection (odds ratio, 10.181; 95% confidence interval, 1.570-66.011, P = 0.015).
Conclusion
The rates of renal dysfunction and pulmonary complications after hepatic resection are higher in patients with LVDD than in those with normal LV diastolic function. Thus, these patients require delicate fluid management.
Despite recent advances in surgical techniques and instruments, blood loss during hepatic resection remains a major concern that significantly influences postoperative morbidity and mortality (Kingham TP et al., 2015). Given the different inflow and outflow systems of the hepatic vasculature, bleeding from the hepatic vein, which directly drains into the inferior vena cava, can be massive and difficult to control (Topaloglu S et al., 2013). Maintaining a low central venous pressure (CVP) during hepatic resection has emerged as an effective strategy to minimize intraoperative blood loss (Hughes MJ et al., 2015; Zhang XL et al., 2015). Hepatic blood congestion, induced by elevated CVP, leads to an incremental increase in transmural pressure and distension of hepatic veins. These veins are consequently torn easily, promoting blood loss at the time of parenchymal transection. Preoperative fluid restriction is one of the most effective and commonly used methods for lowering CVP (Wang CH et al., 2017). After completion of the hepatic parenchymal transection, large amounts of fluid are administered to compensate for the relative hypotension and potential hypoperfusion of abdominal organs during hepatic resection to prevent dysfunction in the postoperative period (Correa-Gallego C et al., 2015; Sear JW, 2005).
Left ventricular diastolic dysfunction (LVDD) is characterized by abnormal myocardial relaxation and filling during diastole and subsequently increased left ventricular (LV) filling pressure (Nagueh SF et al., 2016). Patients with LVDD present with diminished LV compliance that can be intolerable to volume adjustments. Intravenous fluid administration in these patients is associated with an increased risk of fluid overload that is in turn associated with postoperative morbidity and mortality (Jun IJ et al., 2019). Previous studies reported that LVDD may have adverse impacts on renal function and mortality in sepsis patients who require intravenous fluid administration to maintain organ perfusion (Hong JY et al., 2020; Sanfilippo F et al., 2015). The kidneys are encapsulated organs and are thus sensitive to the effects of tissue edema. Accordingly, renal perfusion would be decreased in patients with fluid overload (Payen D et al., 2008). The incidence of major cardiovascular events is also significantly higher after noncardiac surgery in patients with LVDD (Higashi M et al., 2020).
This study aimed to determine the influence of perioperative changes in fluid balance on the postoperative outcomes of hepatic resection in patients with LVDD. We hypothesized that maintaining low CVP during hepatic resection and compensative administration of large amounts of fluid after completion of hepatic parenchymal transection as a strategy for minimizing intraoperative blood loss may have a negative impact on organ function and increase cardiopulmonary morbidity in patients with LVDD who undergo hepatic resection.
This retrospective study evaluated 206 patients who underwent hepatic resection between March 2015 and February 2021 at our hospital. The exclusion criteria were as follows: upper abdominal surgical history (n = 2); combined operation (n = 4); treatment history of hepatic lesion (n = 1); borderline liver function (n = 2); preoperative renal dysfunction (n = 4); and insufficient clinical data (n = 1). Finally, 192 patients were enrolled in the study and stratified into two groups according to LV diastolic function based on echocardiographic assessment using the American Society of Echocardiography (ASE) and the European Association of Cardiovascular Imaging (EACVI) 2016 recommendations (Nagueh SF et al., 2016): LVDD (group A, n = 39), or others including normal LV diastolic function and indeterminate decision (group B, n = 153). Demographics and operative and postoperative outcomes were compared between the groups.
Preoperative transthoracic echocardiography was performed by sonographers or cardiologists using Philips CX50 (Philips Medical Systems, Bothell, WA, USA), and the findings were interpreted by board-certified cardiologists. Parameters were obtained using comprehensive M-mode two-dimensional Doppler echocardiography from the long- and short-axis parasternal views; apical four-chamber, two-chamber, and long-axis views; and subcostal views. LV ejection fraction (LVEF) was calculated using the biplane approach and modified using Simpson’s method. The peak velocity of the transmitral inflow waveform during early (E) and late (A) diastole was recorded at the tip of the valvular leaflets. The peak velocities were measured at early diastole (septal e’ and lateral e’), and the values were averaged (averaged e’) at the septal and lateral mitral annulus. The peak velocity of tricuspid regurgitation flow (TR velocity) was determined using the continuous Doppler method. Left atrial maximum volume at left ventricular end-systole (LAV) was computed using Simpson’s method and was indexed by body surface area (LAVI).
The patients with preserved LVEF were classified according to their LV diastolic function based on echocardiographic assessment using the ASE and the EACVI 2016 recommendations (Nagueh SF et al., 2016): average E/e’ >14; septal e’ velocity < 7 cm/s or lateral e’ velocity < 10 cm/s, TR velocity > 2.8 m/s, and LAVI > 34 mL/m2. Patients with > 50% of these findings were diagnosed with LVDD, while patients with < 50% of these findings were considered to have normal left ventricular diastolic function. The presence of 50% positive findings resulted in an intermediate decision.
Clinico-demographic data, including age, sex, body mass index, presence of diabetes mellitus, hypertension, hepatitis B virus (HBV) or hepatitis C virus (HCV), and diagnosis, were collected. Type of hepatic resection was classified into major resection, defined as resection of three or more segments, or minor resection. Operative data including operative duration, amount of fluids administered, vasopressor usage, requirement of blood transfusion, urine output, and estimated blood loss were investigated. Perioperative laboratory results of liver function (e.g., total bilirubin [TB], international normalized ratio [INR], and albumin) and renal function (e.g., creatinine at admission; operative day; and postoperative days 1, 3, 5, and 7) were analyzed. In addition, all available intake records, composed of oral and parenteral fluids and output data including urine, gastrointestinal losses, and drains, from operative day to postoperative day 7 were collected. Data regarding the nature and incidence of postoperative complications, intensive care unit (ICU) admission, and length of postoperative hospitalization were also collected. Acute kidney injury (AKI) was defined in accordance with the 2012 Kidney Disease Improving Global Outcomes guidelines (Khwaja A, 2012) which had higher predictability than other criteria for assessing prognosis (Lee HA et al., 2022): increase in serum creatinine by ≥ 0.3 mg/dL within 48 hours; increase in serum creatinine to ≥ 1.5 times baseline within 7 days before surgery; or urine volume < 0.5 mL/kg/h for 6 hours. Postoperative liver insufficiency was defined as a peak postoperative TB level of > 7 mg/dL and/or the presence of ascites > 500 mL/day based on a previous study (Mullen JT et al., 2007).
All patients underwent ultrasonography-guided right internal jugular vein catheterization after tracheal intubation in the operating room, and the position of the catheter was determined using a chest radiograph. Electrocardiogram, pulse oximetry, end-tidal carbon dioxide, invasive radial arterial pressure, CVP, and urine output were monitored. Fluid was not administered preoperatively and was restrictively infused after the start of anesthesia, maintaining CVP less than 5 mmHg until the hepatic parenchymal transection was complete. Thereafter, 1,000–1,500 mL/h of crystalloid and colloid solutions were administered in the operating room. To maintain hemodynamic stability, 5 mg ephedrine was administered when the mean arterial pressure decreased below 60 mmHg. Red blood cells were transfused if the hemoglobin concentration decreased to < 8 g/dl in the perioperative period.
All hepatic resections were performed by one surgeon using the same hepatic parenchymal transection technique. The extent of hepatic resection was determined based on tumor size and location. Parenchymal transection was performed using an ultrasonic aspirator, metal clips, and electrocautery device, and the cutting surface of the liver was sprayed with biological glue.
For intergroup comparisons, the data distribution was initially evaluated for normality using the Shapiro–Wilk test. Normally distributed data were presented as means ± standard deviations, and between-group comparisons were conducted using Student’s t-test or Kruskal–Wallis test. Meanwhile, between-group comparisons of descriptive data were conducted using the χ2 test and Fisher’s exact test. Multivariate analysis using an ordinary logistic regression model was performed to investigate the risk factors for specific postoperative morbidities. All statistical analyses were conducted using SPSS Statistics for Windows, version 19.0 (IBM Corp., Armonk, NY, USA).
The median age was significantly higher in group A than in group B (67 [50–87] years vs. 54 [22–82] years, P = 0.000). Hypertension was also significantly more prevalent in group A than in group B (69.2% vs. 37.9%, P = 0.000). Meanwhile, there were no significant differences in sex distribution, body mass index, prevalence of diabetes mellitus, presence of HBV or HCV, diagnosis, and baseline liver function (TB, albumin, and INR) and renal function (creatinine level) between the two groups. Major resections, such as right hemi-hepatectomy, extended right hemi-hepatectomy, left hemi-hepatectomy, extended left hemi-hepatectomy, and central hepatectomy, were more significantly more frequent in group B than in group A (38.5% vs. 63.4%, P = 0.005; Table 1).
| Group A (n = 39) | Group B (n = 153) | P value | |
|---|---|---|---|
| Age (years), median (range) | 67 (50–87) | 54 (22–82) | 0.000 |
| Sex (male) | 27 (69.2%) | 98 (64.1%) | 0.545 |
| Body mass index, kg/m2 | 25.2 ± 3.2 | 25.1 ± 5.7 | 0.957 |
| Diabetes mellitus | 11 (28.2%) | 27 (17.6%) | 0.145 |
| Hypertension | 27 (69.2%) | 57 (37.9%) | 0.000 |
| Presence of HBV | 13 (33.3%) | 39 (25.5%) | 0.337 |
| Presence of HCV | 0 (0%) | 3 (2.0%) | 0.377 |
| Diagnosis | |||
| Hepatocellular carcinoma | 24 (61.5%) | 102 (66.7%) | |
| Cholangiocarcinoma | 4 (10.3%) | 7 (4.6%) | |
| Colorectal liver metastasis | 9 (23.1%) | 28 (18.3%) | |
| Other benign liver disease | 2 (5.1%) | 16 (10.5%) | 0.364 |
| Baseline liver function | |||
| Total bilirubin | 0.7 ± 0.3 | 0.6 ± 0.2 | 0.215 |
| Albumin | 4.2 ± 0.4 | 4.2 ± 0.5 | 0.656 |
| INR | 1.07 ± 0.08 | 1.08 ± 0.09 | 0.363 |
| Baseline renal function | |||
| Creatinine | 0.81 ± 0.27 | 0.74 ± 0.20 | 0.079 |
| Procedure, n (%) | |||
| Major resection (≥ 3 segments) | 15 (38.5%) | 97 (63.4%) | |
| Minor resection (< 3 segments) | 24 (61.5%) | 56 (36.6%) | 0.005 |
| Data are presented as the mean ± standard deviations or n (%) unless otherwise indicated | |||
| Abbreviations: HBV, hepatitis B virus; HCV, hepatitis C virus; INR, International normalized ratio | |||
The mean operative duration was significantly longer in group B than in group A (187.4 ± 79.7 mins vs. 238.3 ± 91.3 mins, P = 0.002). However, there were no significant between-group differences in the intraoperative fluid administration of crystalloid and colloid, requirements of vasopressor or blood transfusion, amount of urine output, and estimated blood loss. In both groups, CVPs showed a decreasing trend after the start of surgery until completion of hepatic parenchymal transection and then re-increased after fluid challenge. CVP at the end of surgery was significantly between the two groups (6.35 ± 3.45 vs. 5.12 ± 2.80, P = 0.025, Fig. 2).
With respect to postoperative complications, the incidence rates of AKI (10.3% vs. 1.3%, P = 0.004) and pleural effusion or edema (51.3% vs. 30.1%, P = 0.013) were significantly higher in group A than in group B. Only one patient had myocardial infarction, and the patient was from group A (2.6% vs. 0%, P = 0.047). Postoperative hepatic insufficiency was more common in group A than in group B (23.1% vs 23.5%), but the difference was not significant (P = 0.953, Fig. 3) There was also no significant difference in the ICU admission rate. No patient died. Postoperative hospital stay was longer in group A than in group B (13.6 ± 6.4 days vs. 12.4 ± 7.0days, P = 0.307), but the difference was not significant (Table 2).
| Group A (n = 39) | Group B (n = 153) | P value | |
|---|---|---|---|
| Operative duration, min | 187.4 (± 79.7) | 238.3 (± 91.3) | 0.002 |
| Intraoperative fluids | 2,330 (± 1,730) | 2,892 (± 1,505) | 0.107 |
| Crystalloid, mL | 2,098 (± 1,289) | 2,536 (± 1,207) | 0.131 |
| Colloid, mL | 232 (± 263) | 356 (± 304) | 0.124 |
| Vasopressor usage | 10 (25.6%) | 35 (22.9%) | 0.748 |
| Blood transfusion | 6 (15.4%) | 24 (15.7%) | 0.936 |
| Urine output, mL | 333 (± 248) | 491 (± 588) | 0.108 |
| Estimated blood loss, mL | 503 (± 343) | 641 (± 796) | 0.297 |
| Intensive care unit admission | 7 (17.9%) | 26 (17.0%) | 0.916 |
| Mortality | 0 | 0 | NS |
| Postoperative hospital stay, days | 13.6 (± 6.4) | 12.4 (± 7.0) | 0.307 |
| Data are presented as the mean ± standard deviations or n (%) unless otherwise indicated | |||
There were no significant between-group differences in TB, INR, and albumin levels as measures of liver function. Creatinine levels were significantly higher in group A than group B from postoperative day 1 to day 7 (Fig. 4).
Comparison of the daily fluid balances from operative day to postoperative day 7 to identify the differences in fluid management showed no significant differences between the two groups. However, daily urine outputs were significantly lower in group A than in group B at postoperative day 1 (1,560 ± 577 vs. 1,853 ± 928, P = 0.038) and day 2 (1,744 ± 971 vs. 2,036 ± 853, P = 0.024; Fig. 5).
LVDD was the only significant predictor of AKI in univariate analysis (odds ratio [OR], 8.629; 95% confidence interval [CI], 1.519–49.000, P = 0.015) and multivariate analysis (OR, 10.181; 95% CI, 1.570–66.011, P = 0.015; Table 3).
| Univariate analysis | Multivariate analysis | |||||
|---|---|---|---|---|---|---|
| Variable | OR | 95% CI | P value | OR | 95% CI | P value |
| Age (year) | 1.020 | 0.966–1.076 | 0.485 | |||
| Sex (male) | 0.360 | 0.315–24.036 | 0.360 | |||
| Body mass index | 1.067 | 0.986–1.154 | 0.108 | |||
| Diabetes mellitus | 4.314 | 0.835–22.285 | 0.081 | |||
| Hypertension | 2.593 | 0.463–14.507 | 0.278 | |||
| Presence of HBV | 1.360 | 0.242–7.656 | 0.727 | |||
| Presence of HCV | NS | |||||
| Total bilirubin | 2.653 | 0.121–58.212 | 0.536 | |||
| INR | 2.029 | 0.000-8457.360 | 0.868 | |||
| Albumin | 0.666 | 0.112–3.948 | 0.654 | |||
| Creatinine | 0.239 | 0.003–19.011 | 0.522 | |||
| Operation time (min) | 1.002 | 0.994–1.011 | 0.596 | |||
| Major resection | 0.706 | 0.139–3.594 | 0.675 | |||
| LVDD | 8.629 | 1.519-49.000 | 0.015 | 10.181 | 1.570-66.011 | 0.015 |
| Vasopressor usage | 0.618 | 0.070–5.435 | 0.665 | |||
| Transfusion | 1.041 | 0.117–9.245 | 0.971 | |||
| Estimated blood loss | 1.000 | 0.999–1.001 | 0.807 | |||
| Abbreviations: HBV, hepatitis B virus; HCV, hepatitis C virus; INR, International normalized ratio; LVDD, left ventricular diastolic dysfunction; OR, odds ratio; CI, confidence interval | ||||||
This retrospective study showed that the rates of renal dysfunction and pulmonary complications after hepatic resection are higher in patients with LVDD than in those with normal LV diastolic function.
LVDD as the most common cardiac abnormality in patients with liver cirrhosis is characterized by abnormal myocardial relaxation related to the development of myocardial fibrosis, hypertrophy, and subendothelial edema (Fouad YM et al., 2014). It occurs when the passive elastic properties of the myocardium are reduced because of increased myocardial mass and changes in extracellular collagen (Aurigemma GP et al., 2004). LVDD is prevalent in 25.7–81.4% of cirrhosis patients; it is also more prevalent in decompensated cirrhosis patients than in compensated cirrhosis patients (Farouk H et al., 2017). Liver cirrhosis, old age, hypertension, diabetes, and LV hypertrophy are major risk factors of LVDD (Jeong EM et al., 2015). In the current study, 37/190 patients (19.5%) had LVDD, and these patients were older and had a higher prevalence of hypertension than those with normal LV diastolic function.
Maintaining a CVP during hepatic resection and compensative administration of large amounts of fluid after its completion is a widely used effective strategy to minimize intraoperative blood loss. A previous study showed that clinically relevant renal dysfunction is highly uncommon in patients with a low CVP who undergo hepatic resection (Correa-Gallego C et al., 2015). However, patients with LVDD are vulnerable to the changes in fluid status. Lowering preload during the operation could decrease cardiac output, and it might lead to hypoperfusion of abdominal organs, making it a risk factor for postoperative renal dysfunction. Analysis for intraoperative changes of cardiac stroke volume would be helpful to reveal its influence on clinical outcomes, but this study was a retrospective nature and it might be a limitation. On the other hand, patients with LVDD are also at risk of fluid overload that fluid challenge after completion of hepatic resection can conversely cause postoperative renal dysfunction. Fluid overload is associated with a high risk of AKI and delayed recovery because tissue edema of the kidneys leads to hypoperfusion-induced organ injury (de Oliveira FS et al., 2015). Previous studies have shown that patients with LVDD often have elevated CVP; this in turn has a negative correlation with estimated glomerular filtration rate and increases the risk of postoperative AKI (Damman K et al., 2009; Palomba H et al., 2007). In the current study, CVP at the end of surgery with fluid challenge after hepatic parenchymal transection was higher in patients with LVDD. The incidence of postoperative AKI was also significantly higher in these patients than in those with normal LV diastolic function despite a similar fluid balance between them. In addition, we used CVP to assess the volume status although it has limited accuracy (De Backer D et al., 2018). Further studies to confirm this results are required using more reliable indicators reflecting fluid status and responsiveness, such as inferior vena cava size.
The current study showed that patients with LVDD had significantly lower daily urine outputs in the early postoperative period than those with normal LV diastolic function. Fluid therapy in this period would be important to avoid fluid overload, which is associated with the development of AKI. We adjusted the amounts of postoperative fluid administration in accordance with urine outputs so that there were no significant differences in daily fluid balances between the two groups. This might have influenced the result of no significant difference in postoperative hospital stay between patients with and without LVDD. All cases of postoperative AKIs were resolved at discharge without requiring renal replacement therapy. However, AKI was associated with prolonged postoperative hospital stay (22.0 ± 14.3 days vs. 12.3 ± 6.3 days, P = 0.001). Thus, usage of nephrotoxic medications, particularly those that cause either glomerular or interstitial damage, should be avoided in patients with LVDD, especially in the early postoperative period.
LVDD is associated with adverse postoperative cardiovascular events (Fayad A et al., 2016). Previous studies showed that LVDD is an independent risk factor for postoperative pulmonary edema in patients undergoing non-cardiac surgery owing to the increase in LV filling pressure concomitant with pulmonary venous pressure (Shigematsu K et al., 2019). The current study also found that pleural effusion or pulmonary edema was significantly higher in patients with LVDD than in those with normal diastolic function. Although most cases of pleural effusion or pulmonary edema were spontaneously improved with or without diuretics, two patients with LVDD (5.1%) required percutaneous drainage because of increasing oxygen requirement. LVDD can also impair coronary flow reserve, increasing LV wall stress in patients with normal diastolic function and leading to increased myocardial blood flow (Galderisi Met al., 2002). Patients with LVDD are at risk of coronary flow reduction after volume adjustment, and this would result in myocardial ischemia and adverse cardiovascular events. In the current study, one patient with LVDD had myocardial infarction requiring a stent placement at postoperative day 1.
There are some limitations to this study. The retrospective study design required us to rely on the integrity of completed medical records for our analysis. In addition, the study population was relatively small, and thus, we only compared the clinical outcomes between patients with and without LVDD. However, the outcomes differ by grade of LVDD, and thus, further large-scale studies with subgroup analysis by grade of LVD are needed. Finally, additional prospective studies are needed to confirm that fluid therapy for preventing fluid overload in the perioperative period would improve the clinical outcomes of patients with LVDD.
Patients with LVDD have significantly higher CVP after fluid challenge at the end of surgery compared to those with normal LV diastolic function. This high CVP might be associated with increased rates of postoperative AKI and pleural effusion or edema because of fluid overload. Despite similar daily fluid balances, renal dysfunction as indicated by increases in serum creatinine and decreases in daily urine outputs is more common in patients with LVDD than in those with normal LV diastolic function. Further, fluid overload might be related with the development of postoperative morbidities in patients with LVDD. Thus, fluid therapy in the perioperative period of hepatic resection is important for patients with LVDD at risk of fluid overload.
Acknowledgements
Not applicable
Authors’ contributions
SSW is the principal investigator of this study. SJ collected the data and assisted in the statistical analysis, outlined the analysis to be performed, and made figures. SSW conceived the idea, collected the data and performed a statistical analysis, outlined the analysis to be performed, wrote the manuscript and critically reviewed and edited the manuscript. All authors read and approved the final manuscript.
Funding
None
Availability of data and materials
The datasets used and analyzed during the current study are available from
the corresponding author on reasonable request.
Ethics approval and consent to participate
This study was approved by our Institutional Review Board (IRB No. 2108-009-19379) and performed in accordance with the Declaration of Helsinki (as revised in 2013). The need for informed consent was waived because accrual patient records were analyzed, and no patient identification data were used.
Consent for publication
Not applicable
Competing interests
The authors declare that they have no competing interests.